• Bulletin of the Korean Chemical Society
• /
• v.28 no.9
• /
• pp.1472-1476
• /
• 2007

Artificial neural networks (ANNs) are successfully developed for the modeling and prediction of normalized polarity parameter (ETN) of 216 various solvents with diverse chemical structures using a quantitative-structure property relationship. ANN with architecture 5-9-1 is generated using five molecular descriptors appearing in the multi-parameter linear regression (MLR) model. The most positive charge of a hydrogen atom (q+), total charge in molecule (qt), molecular volume of solvent (Vm), dipole moment (μ) and polarizability term (πI) are input descriptors and its output is ETN. It is found that properly selected and trained neural network with 192 solvents could fairly represent the dependence of normalized polarity parameter on molecular descriptors. For evaluation of the predictive power of the generated ANN, an optimized network is applied for prediction of the ETN values of 24 solvents in the prediction set, which are not used in the optimization procedure. Correlation coefficient (R) and root mean square error (RMSE) of 0.903 and 0.0887 for prediction set by MLR model should be compared with the values of 0.985 and 0.0375 by ANN model. These improvements are due to the fact that the ETN of solvents shows non-linear correlations with the molecular descriptors.

• Journal of Korean Society of Environmental Engineers
• /
• v.35 no.12
• /
• pp.897-905
• /
• 2013

In this study, sorption coefficients (${\log}K_{OC}$, n) for the binding of phenanthrene (PHE) to soil humins, insoluble fraction of soil humc substances (HS), were determined and relationship between the sorption coefficients and structural characteristics of the soil humins were investigated. The soil humins used in the present study were isolated from 7 different soils including 5 domestic soils, an IHSS standard and a peat soil, and characterized by elemental analysis and CPMAS $^{13}C$ NMR method. $^{13}C$ NMR spectral features indicate that the soil humins are mainly made up of aliphatic carbons (57.1~72.3% in total carbon) with high alkyl-C moiety, and the alkyl-C contents ($C_{Al-H,C}$, %) was in order of granite soil Hu (26~42%) > volcanic ash soil, HL Hu (23.9%) > Peat Hu (14.0%). The results of correlation study show that a positive relationship ($r^2$ = 0.77, p < 0.05) between organic carbon normalized-sorption coefficients ($K_{OC}$, mL/g) and alkyl-C contents($C_{Al-H,C}$, %), while negative relationship ($r^2$ = (-)0.74, p < 0.05) between Freundlich sorption parameter (n) and H,C-substituted aromatic carbon contents ($C_{Ar-H,C}$, %). The magnitude of $K_{OC}$ values are also negatively well correlated with polarity index (e.g., PI, N + O)/C) ($r^2$ = (-)0.74, p < 0.1). These results suggest that the binding capacity (e.g., $K_{OC}$) for PHE is increased in soil humin molecules having high contents of alkyl-C or lower polarity, and nonlinear sorption for PHE increased as the H,C-substituted aromatic carbon contents ($C_{Ar-H,C}$, %) in the soil humins increased. The PHE sorption characteristics on soil humins are discussed based on the dual reactive mode of sorption model.